As the human and economic toll of HIV/AIDS becomes increasingly apparent, governments and international agencies are initiating prevention and treatment programs. While the medical effectiveness of antiretroviral therapy has been well described, little is known about the economic effectiveness of large-scale provision of treatment. The extent to which antiretroviral therapy can mitigate the impacts of AIDS on labor productivity, household welfare, and business competitiveness is unknown. The overall goal of the proposed study is to estimate the labor productivity consequences of providing antiretroviral therapy to agricultural workers in sub-Saharan Africa. The study has four specific aims: 1) estimate the impact of antiretroviral therapy on the labor [unreadable] productivity of HIV-positive workers; 2) describe the relationship between biomedical and economic indicators of disease progression; 3) assess the impact of treatment on household welfare and firm operating costs; and 4) strengthen the capacity of our Kenyan collaborating institution. The proposed, prospective cohort study will use data on HIV-positive workers on agricultural estates in western Kenya who enroll in HIV/AIDS treatment programs. Three data sets will be collected: biomedical indicators (e.g. disease stage and progression, CD4 and treatment regimen); productivity indicators collected by [unreadable] agricultural estates (e.g. daily job attendance and weight of crop harvested daily); and self-reported data about morbidity, income, and labor substitution. The study will estimate the effect of treatment on three aspects of productivity-impaired performance, absenteeism, and shifts to light tasks-and on worker household welfare and employer production costs. Approximately 450 index subjects on treatment will be matched to reference subjects who are either known to be HIV-negative, are of unknown HIV status, or previously died of AIDS-related causes. By adding this analysis to an existing public health intervention, the study offers the opportunity to link high quality medical data with detailed, individual data on labor productivity and household welfare. The study design overcomes many of the limitations of previous work on health and productivity, while filling a critical gap in knowledge about HIV/AIDS. The information generated will assist policy makers and donor agencies throughout the developing world to design well targeted, economically efficient, and sustainable responses to the AIDS [unreadable] epidemic. [unreadable] [unreadable]